Exploring the role of ChatGPT in medical image analysis

Abstract

The rapid advancement of artificial intelligence (AI) in the field of medical image analysis has demonstrated various applications such as computer-aided disease diagnosis and prognosis, image registration, tissue segmentation, image fusion, annotations, etc. However, its real-time use in routine clinical settings is still far-fetched due to numerous factors including lack of clinical trials in AI, development of diverse datasets, generalization of AI models in real-time, their interpretability, and associated biases. The introduction of Chat Generative Pre-trained Transformer (ChatGPT) has further created both an opportunity and havoc in the minds of researchers for its use and applications in this field. This editorial aims to highlight the role of ChatGPT in the field of medical image analysis. An exploratory analysis was conducted on ChatGPT by asking more than a hundred questions related to this field like “what is medical image analysis”, “how can you help me with medical image analysis”, “generate a code for --- to perform medical image analysis”, “perform literature survey on --- for medical datasets”, “analyze the anomaly in this frame”, “write an editorial on ---”, “give road map for --- in medical image analysis” etc in February and May 2023. The inferences of the responses generated by ChatGPT, their variation over a two-month gap, our viewpoint, its shortcomings, and concluding remarks have been discussed in this editorial.